Erin Servais, AI for Editors
August 15, 2025
Technical communicators face a constant challenge: delivering the right information to the right audience without drowning in duplicate work.
The traditional “one‑size‑fits‑all” approach rarely works. Content for beginners overwhelms with detail, while advanced content alienates new users. Meeting diverse needs often requires creating multiple versions of the same documentation, a time‑intensive process few teams can sustain.
At ConVEx 2025, I demonstrated a practical solution that blends AI efficiency with content strategy: custom AI assistants, such as ChatGPT’s Custom GPTs and Google Gemini’s Gems. Think of these tools as mini-AIs with a single mission. In the use case I demoed, they can adapt your content automatically by adjusting complexity, format, or style while maintaining accuracy and consistency.
And yes, in my demo, there was a time machine.
Why Technical Communicators Should Pay Attention
Documentation teams are under pressure to produce more content in more formats, often with flat or shrinking budgets. Custom AIs can close this gap by providing:
- Efficiency – AI can generate multiple content versions in seconds, cutting down on repetitive writing tasks.
- Consistency – A well‑designed custom AI follows defined patterns and style rules, ensuring outputs match your team’s voice.
- Adaptability – AI instructions and uploaded knowledge documents can be updated for new product releases or feature changes, making these tools future‑proof.
With custom AIs, even tiny but mighty teams can have an outsized impact, automating tasks that once required large‑scale content ops or IT support.
How to Build a Custom AI Using the 4-P Framework
A custom AI assistant is essentially a teammate that never tires. You “teach” it with instructions, and it follows them, repeatedly and reliably.
I teach teams to use the 4-P Framework when writing the assistant’s instructions. This formula is simple but effective and includes the essential elements to build powerful assistants.
The custom AI I built to adapt and multiply content is called Tech Writing Leveler (URL: https://chatgpt.com/g/g-fAcyhZS6Z-tech-writing-leveler). Let’s walk through its 4 Ps.
Persona – Assign a role.
You are an expert technical writer who specializes in writing content for readers with varying knowledge levels.- Purpose – Define its task.
Your purpose is to assess technical writing user inputs and produce three versions of that content: one version suitable for a beginner, one suitable for a person with intermediate knowledge of the subject, and one for an expert. - Process – Outline the steps.
Analyze the user input. Then create a beginner-level version, intermediate-level version, and an expert-level version of the content. - Pattern – Explain how it should respond. It helps to include brief examples.Follow this example output structure:
Beginner Version: DITA (Darwin Information Typing Architecture) is a way of organizing and writing technical content.
Intermediate Version: DITA (Darwin Information Typing Architecture) is an XML-based standard for structured content creation and management.
Expert Version: DITA (Darwin Information Typing Architecture) is an open XML-based standard for structured content authoring, emphasizing modularity, reusability, and semantic tagging.
When paired with knowledge files (style guides, glossaries, or sample docs with writing formats and styles for it to follow), these instructions turn a generic AI into a repeatable workflow engine.
Case Study: The Chrono‑Chamber 5000
To illustrate the potential of AI‑automated content adaptation, I created an as-yet-fictional product: the Chrono‑Chamber 5000, a time‑travel device with complex technical documentation.
The challenge: Users range from casual adventurers in the space-time continuum to seasoned quantum physicists. Traditional documentation would require separate manuals or a frustrating compromise between clarity and completeness.
The solution: My Tech Writing Leveler GPT took the original dense documentation and, using the 4-P Framework’s instructions, transformed it into audience‑specific versions.
You can see the complete set of three documentation versions here (URL: https://chatgpt.com/share/67de6529-7954-8007-9a86-f8946e24153f). Let’s pull out one definition to get a quick look at what the tool can do.
Original:
Micro-Singularity Generator: At the heart of the chamber lies a micro-singularity generator. This device creates a controlled, temporary Kerr-Newman singularity (a rotating, charged black hole) with an extremely small event horizon. This singularity acts as the “engine” for time travel.
Beginner:
Mini Black Hole Generator: A tiny, controlled black hole powers the chamber, helping to create the time tunnel.
Intermediate:
Micro-Singularity Generator: A controlled Kerr-Newman singularity (a charged, rotating black hole) acts as the propulsion mechanism for time travel.
Expert:
Micro-Singularity Generator: A high-stability Kerr-Newman singularity serves as the primary mechanism for creating traversable temporal conduits.
This demonstration shows AI’s practical, real‑world time savings. Producing three complete product guides manually could take hours, while the AI generates them in under a minute.
Next-Level Use: Multi‑Format Adaptation
Custom AIs can also transform content into different formats, enabling technical writers and editors to scale one piece of content to meet more needs.
In my demo, I showed how the Tech Writing Leveler could turn a section of the Chrono‑Chamber 5000’s manual into an FAQ, while still adjusting for the three knowledge levels (URL: https://chatgpt.com/share/e/688fed98-d690-8001-a1fb-b930faed72f4). Here is an excerpt:
Original:
Paradox Prevention System: While the TNS incorporates a sophisticated paradox prevention system, it is not foolproof. Users are strongly advised to avoid actions that could significantly alter established historical events.
Beginner:
Q: Can I change the past without causing problems?
A: The machine has a system to prevent time paradoxes, but it’s not perfect. Avoid actions that could change big events in history.
Intermediate:
Q: How reliable is the paradox prevention system?
A: The integrated Temporal Navigation System (TNS) features a paradox prevention mechanism, but it is not infallible. Users should avoid interactions that may significantly alter recorded history.
Expert:
Q: How effective is the TNS in mitigating paradoxes?
A: While the Temporal Navigation System (TNS) incorporates an advanced paradox prevention protocol, it remains probabilistic rather than absolute. Users are urged to minimize interventions with established historical events to avoid temporal destabilization.
Suddenly, a wall of intimidating text becomes a user‑friendly reference. This ability to reformat and level content simultaneously creates huge downstream efficiencies. Teams can quickly spin up quick‑start guides, video scripts, chatbots, or FAQs from a single canonical source document (about time travel or any other subject).
Practical Takeaways for Documentation Teams
If you want to explore AI‑driven content adaptation, here’s a roadmap:
- Identify a high‑impact, low‑risk task.
Examples: multi‑audience procedures, glossary generation, or summaries. - Develop your assistant’s instructions.
Use the 4 Ps to ensure quality, repeatable outputs. - Test and refine.
Start small, verify outputs with human review, and iterate on instructions. - Add knowledge files for accuracy.
Upload your style guides, brand guides, and sample documents. - Scale thoughtfully.
Once the workflow proves itself, expand to new formats, products, or audiences.
Personalization Without the Pain
The next phase of technical communication will hinge on content adaptability. Users will expect documentation that feels tailored to their knowledge level, device, and use case, without waiting for multiple documentation cycles.
Teams that will succeed will view AI not as a replacement but as a force multiplier, a digital teammate that handles first drafts and repetitive adaptation so they can stay on top of workloads and serve their customers better. By augmenting workflows with custom AIs like Custom GPTs and Gemini Gems, these teams will deliver that personalization without increasing their effort.
And if your current content projects feel overwhelming, remember the lesson of the Chrono‑Chamber 5000: sometimes, the fastest path to the future is to let AI open the portal.
Erin Servais teaches writers, editors, and content teams to upskill with AI through her AI for Editors courses and custom training programs. To learn how she can help your team, visit https://www.aiforeditors.com/ and https://www.servaissolutions.com/ or email her at [email protected].
This article is based on Erin Servais’s original presentation notes and slides, which were transformed into a draft using OpenAI’s ChatGPT-4o. Erin Servais then comprehensively and substantially edited the AI-generated draft to produce the final version and meet standards for quality and authenticity. The human author retains full responsibility for the final content and its adherence to ethical guidelines.